google-deepmind/robotics-antigravity-plugin

i2rt-yam-bringup

Checks and installs the I2RT YAM SDK (i2rt), configures Linux socketcan or macOS gsusb CAN bringup for I2RT YAM robot arms, and provides a wave smoke test (wave.py).

查看源码

使用方法

  1. 复制安装命令,在自己的智能体环境安装前先检查源码包。
  2. 从原始文档核对所需工具、凭据和支持的宿主。
  3. 说明输入材料与期望交付物,并对照来源检查输出。
npx skills add https://github.com/google-deepmind/robotics-antigravity-plugin --skill i2rt-yam-bringup

开始前需了解

  • 安装不包含模型订阅、API 额度或访问文件的权限。
  • 仓库星数和安装量反映活跃情况,不等于独立质量或安全测试。
仓库原始内容

按源仓库内容呈现,保留标题、案例、代码、表格、链接以及原文引用的演示图片。

I2RT YAM SDK & CAN Bringup

Ensure the official I2RT YAM Python SDK (i2rt) and host CAN transport are ready before running YAM robot code (i2rt drives physical YAM Damiao motors over 1 Mbps CAN via python-can; only get_yam_robot(..., sim=True) runs without CAN):

  1. SDK Check & Install:
  • Check python3 -c "from i2rt.robots.utils import ArmType, GripperType; print('i2rt ready')".
  • If missing (i2rt is not on PyPI), clone git clone https://github.com/i2rt-robotics/i2rt.git and install with pip install -e ./i2rt (or pip install --no-deps --no-build-isolation -e ./i2rt plus numpy mujoco mink ruckig python-can pyyaml portal crcmod-plus dm-env if pyproject.toml pins conflict with host Python).
  1. Real-Hardware CAN Bringup (Linux `socketcan` vs. macOS [`macos_can.py`](scripts/macos_can.py)):
  • Linux: Bring up socketcan at 1 Mbps (sudo ip link set can0 up type can bitrate 1000000) and pass channel="can0" to get_yam_robot.
  • macOS (Apple Silicon & Intel): i2rt's get_yam_robot and run_startup_checks require "can" in channel and hardcode bustype="socketcan", which macOS lacks; moreover, python-can's built-in GsUsbBus calls detach_kernel_driver(0) (Access denied without sudo) and dev.reset() (wedging gs_usb STM32 EP 0x81 when TX echoes are unread).
  • On macOS, ensure libusb and gs-usb are installed (brew install libusb && pip install gs-usb pyusb for candleLight / CANable / I2RT USB-CAN adapters) and call install_mac_can() from bundled `macos_can.py` once before get_yam_robot(channel="can0", ...):
  • Safe no-op on Linux; auto-discovers libusb-1.0.dylib across /opt/homebrew/lib, /usr/local/lib, and /opt/local/lib inside isolated venv/conda environments without DYLD_LIBRARY_PATH.
  • Maps "can0", "can1", ... deterministically across single or bimanual gs_usb adapters without detach_kernel_driver or dev.reset(), filters TX echoes (echo_id != GS_USB_NONE_ECHO_ID), drains clean_error bursts, runs a silent motor-independent GS_CAN_MODE_LOOP_BACK health check (with DFU auto-recovery only if EP 0x81 is wedged), and also supports PCAN (interface="pcan") or SLCAN (interface="slcan", backend_channel="/dev/cu.usbmodem...").
  1. Safe Shutdown:
  • Park the arm gently near the table before calling robot.close() so it never drops (since robot.close() sets all motor torques to zero).
  1. Bring-Up Wave Task ([`wave.md`](references/tasks/wave.md), [`wave.py`](scripts/wave.py)):
  • Run `wave.py` (--sim offline or --channel can0 on

hardware) for a free-space wave smoke test; see `references/tasks/wave.md` for keyframes and flags.

来自同一仓库

更多 Skills

全部 Skills
google-deepmind
社区

blender-creation

Creates physically calibrated, articulated, PBR-textured 3D MuJoCo assets (MJCF model.xml, .mjz, mjSpec) from scratch using headless Blender (bpy, bmesh, blendermeshutils, assetbuildutils), convex collision synthesis, 10-scenario auditing, and Studio preview. Use only when asked to create, model, or build a 3D MuJoCo object from a URL, image, or text description, or when blender-creation is named. Don't use for non-MuJoCo Blender/Cycles/CAD modeling.

1小时安装量
2
GitHub Stars
80
最近更新
11月17日
google-deepmind
社区

fr3-duo-control

Franka Duo robot only (Franka Research 3 / FR3 Duo dual-arm cell with Robotiq 2F-85 grippers, ZED Mini and RealSense D405 cameras). Brings up, homes, and controls it via pylibfranka on hardware or in MuJoCo sim. Use only when the user names Franka Duo, FR3 Duo, or fr3duo.

1小时安装量
2
GitHub Stars
80
最近更新
11月17日
google-deepmind
社区

i2rt-yam-ik

Solves 6-DOF inverse kinematics for the I2RT YAM and YAM Pro robot arm using Mink + DAQP on combined YAM + gripper MJCF. Use only when computing joint targets or local Cartesian moves on an I2RT YAM or YAM Pro robot arm. Don't use for non-YAM inverse kinematics, general quadratic programming (QP) solvers, or non-robotics tasks.

1小时安装量
2
GitHub Stars
80
最近更新
11月17日
google-deepmind
社区

iteration

Staged execution and video-based debugging workflow for physical or MuJoCo simulated robots: splitting a robot manipulation task into stages, testing each stage with one script, reviewing rollout videos, and verifying physical completion after release and settle. Use only when writing code that makes a physical or simulated robot act and move. Don't use for non-robotics software engineering, UI/WebGL testing, or general scripting.

1小时安装量
2
GitHub Stars
80
最近更新
11月17日